Curiosity-Driven Development of Action and Language in Robots Through Self-Exploration
This research addresses the problem of inefficient learning in AI compared to human infants, offering insights into developmental mechanisms, though it is incremental in applying active inference and reinforcement learning to robotics.
The study tackled how robots can efficiently learn actions and language through self-exploration, finding that curiosity-driven exploration with motor noise substantially outperforms learning without curiosity and that generalization improves as compositional elements increase.
Human infants acquire language and action gradually through development, achieving remarkable generalization capabilities from only a minimal number of learning examples. In contrast, recent large language models require exposure to billions of training tokens to achieve such generalization. What mechanisms underlie such efficient developmental learning in humans? This study addresses this question through simulation experiments in which robots learn to perform various actions corresponding to imperative sentences (e.g., \textit{push red cube}) via trials of self-guided exploration. Our approach integrates the active inference framework with reinforcement learning, enabling curiosity-driven developmental learning. The simulations yielded several important findings: i) Generalization is drastically improved as the number of compositional elements increases. ii) Curiosity-driven exploration combined with motor noise substantially outperforms learning without curiosity. iii) Rote pairing of sentences and actions occurs before the emergence of compositional generalization. iv) Simpler, prerequisite-like actions emerge earlier in development, while more complex actions involving these prerequisites develop later. These results shed light into possible mechanisms underlying efficient developmental learning in infants and provide computational parallels to findings in developmental psychology.